Three recent empirical studies on the use of placebos and two papers arguing for the deliberate use of placebos in clinical practice are analysed. Empirical studies demonstrate that placebos are commonly used. The concept of the placebo is currently understood in different ways, many of which do not refer to inert substances or treatments. The papers arguing for the use of placebos are shown to fail to make their case.
Statistics from Altmetric.com
Doctors have lied to their patients throughout history. Mentiris ut medicus (“lie like a doctor”) was a popular proverb in the Middle Ages,1 and in the 17th century Bishop Jeremy Taylor encouraged physicians to lie if it helped the patients.2 Lying was, however, to be understood with charity and with honour to the profession.
Telling the truth as a moral imperative is, in fact, a fairly recent phenomenon and is closely related to the development of patient autonomy in the Western world since the 1960s. It is worth noting also that withholding the truth—a serious diagnosis, for example—is still common practice in many, maybe even most, parts of the world.
Deliberate use of a placebo, defined as a treatment without any effect as such, is usually considered unethical. The Council on Ethical and Judicial Affairs of the American Medical Association, for example, states that it “may undermine trust, compromise the patient–physician relationship, and result in medical harm to the patient”.3
Physicians’ attitudes on the use of a placebo vary broadly, however. Examples of two extreme positions are those of Brown (“As physicians, we should respect the benefits of placebos—their safety, effectiveness and low cost—and bring the full advantage of these benefits into our everyday practices”)4 and Hróbjartsson (“Clinical placebo interventions are unethical, unnecessary, and unprofessional”).5
Two recent papers in this journal have argued for the deliberate use of placebos.6 7 In the following discussion, I first review recent empirical studies on clinical placebo use. I then scrutinise the two papers and show that both fail to make their case.
STUDIES ON THE USE OF THE PLACEBO
During the past few years, several studies have demonstrated that the deliberate use of placebos is a common and widely accepted practice. A closer look at the three most recent of these studies shows, however, that the placebo concept is understood in several different ways and that the crucial difference between placebo use and placebo effect is often not understood.8–10
A study in Denmark found that 86% of the general practitioners included in the sample reported having used a placebo intervention at least once during the previous year.8 The researchers deliberately used a narrow definition for the placebo, as an intervention not considered to have any “specific” effect on the condition treated but with a possible “unspecific” effect. In the paper, the authors deliberated on the concepts “specific” and “unspecific” as follows: “Whereas this distinction is relatively clear in connection with pharmacological interventions, it is less obvious in connection with nonpharmacological treatments such as physiotherapy.” To me this distinction is not obvious. Although common in the placebo literature, the concepts are vague. What else could “unspecific” refer to than the fact that we do not know the mechanism of an effect?
One typical placebo intervention among Danish physicians concerned antibiotics for viral infections. This use may sound simple, but the interpretation is not clear, as the authors also note in their discussion. Respiratory infections are viral by origin in most cases, but, particularly in general practice, laboratory tests are hardly ever made for confirmation. In prolonged or atypical cases, physicians often prescribe antibiotics although no firm evidence of a bacterial origin exists. Furthermore, many patients have both viruses and bacteria, and it is not even possible to determine for sure whether antibiotics are necessary.
The main finding of a questionnaire survey in Israel was that 60% of the respondents (53% of the physicians and 71% of the nurses sent the questionnaire) reported using a placebo.9 Most of the users found placebos generally or occasionally effective. The results are interesting but are difficult or impossible to interpret because the authors did not specify what they meant by placebo.
Because the key concept was not defined, the respondents may have understood placebo in at least four different ways. First, placebo may have meant deliberate deception through the administration of an inert substance. Second, it may have meant giving an inert substance openly. In fact, 29% of the respondents thought that patients should be informed that they were receiving a placebo. Third, some may have thought of a situation in which the doctor or nurse believes that the drug works even though there is no supporting scientific evidence. Fourth, some respondents may have thought more about the placebo effect than the nature of the substance given.
The respondents’ description of the circumstances of the use of a placebo demonstrates the variation in the understanding. For example, paracetamol and vitamin C were mentioned as examples of placebos given. Although the conditions in which the placebo were used were not reported, it is obvious that neither substance can be labelled inert.
In the abstract of the paper, the authors concluded that “used wisely, placebos might have a legitimate place in therapeutics”. This conclusion is unjustified for two reasons. First, on the basis of the survey it is far from clear what might have a legitimate place in therapeutics. Second, the authors report the results of an empirical descriptive study and do not scrutinise the normative arguments for or against the clinical use of a placebo.
According to a questionnaire survey among internists in Chicago, 45% of the respondents had used placebos in clinical practice.10 The researchers provided several alternatives for the definition of placebo. The respondents could give their own definition or choose between the following alternatives, one of which was exactly the same as in the Danish study: (1) an intervention that is not expected to have an effect through a known physiologic mechanism; (2) an intervention not considered to have a “specific” effect on the condition treated, but with a possible “unspecific” effect; and (3) an intervention that is inert or innocuous.
The definitions demonstrate the conceptual ambiguity related to the placebo. The first definition includes the hint of a possible psychological mechanism. The second contains the problematic terms “specific” and “unspecific”, the problems of which were discussed earlier. The third definition is narrow and refers to the characteristics of the intervention only.
In this study, too, the description of the treatments that were understood as placebos demonstrates the multiple meanings of the term. Antibiotics (for viral or other non-bacterial diagnoses), vitamins, ibuprofen, herbal supplements, and prepared placebo tablets are examples of placebo therapies mentioned by the respondents.
The authors briefly refer to the discussion on the ethics of deliberate placebo use. However, they confuse the use of placebo and placebo effect, when they write: “In the broader ethics literature, some commentators on informed consent and nondeceptive therapeutics caution against the use of placebos in medical practice. Others propose that the placebo effect can be harnessed in various therapeutic contexts that do not pose ethical dilemmas.”
In summary, empirical research shows that the use of placebos is common in clinical practice. However, it also shows that the concept of the placebo is understood in many different ways, and in most cases clinicians refer to substances or treatments that are not inert.
A MORAL IMPERATIVE?
The empirical study in Israel was a starting point for a paper discussing the normative issues related to the clinical use of placebos.6 After briefly introducing the conceptual problems related to placebos, the authors presented three clinical cases on the basis of which they propose guidelines for the clinical use of placebos. They also conclude that, in select cases, the use of placebos may even be morally imperative. The paper contains, however, questionable reasoning and erroneous assumptions.
First, the operational definition of the placebo concept is problematic. Lichtenberg and colleagues write that they
… address the ethics of the placebo operationally by asking when it is ethical, in clinical practice, to offer a pill or perform a procedure as an alternative to, or in the absence of, a standard, proven therapy when the effect, if any, of that pill or procedure is expected to be mediated by psychophysiological mechanisms, such as expectation, relaxation, or conditioned response, or what has elsewhere been termed a “meaning response”. The pill or procedure would then be considered the placebo; the effect it produces would be the placebo effect.6
If a standard, proven therapy exists, how could it ever be ethically acceptable to offer a placebo as an alternative to it? In the absence of a standard, proven therapy, giving a placebo and telling the patient openly about it can be an option. The key moral issue is deception.
Second, the authors connect the morality of placebo prescription to a particular world view: “The placebo is a deception only for those who would reduce treatment to a purely biomedical pursuit.” This statement is not true. An inert treatment as such is not deception, but the act of providing the treatment may or may not contain deception.
Third, only the first of the three cases presented in the paper supports the possible open use of a supposed placebo medication. A 45-year-old man had severe postoperative pain after leg amputation. Opioids did not help and he was offered intramuscular saline. The staff explained that saline had been used as an effective pain killer, and it did help this patient, too.
In the second case, the wife of a patient demanded a shot of penicillin for her husband, who had gastroenteritis. The doctor did not agree to give the shot, but the authors concluded, “Had the family continued to demand treatment beyond reassurance, the doctor might have considered giving saline, admitting it was saline, and assuring the patient that he would rapidly recover.” The authors do not, however, present any arguments to support their claim. A “demand beyond reassurance” is a challenge for communication, not a legitimation for an injection of intramuscular saline.
The third case was that of a 32-year-old woman with depression, who, after failed attempts with hypnotherapy, demanded medication. The psychiatrist prescribed imipramine at a starting dose of 25 mg, explaining that effectiveness generally requires 2 to 4 weeks at a much higher dose. The patient reported remarkable improvement, however, already the next day. The authors thought that imipramine was used here as a placebo. This is not obvious, however. First, even a low dose of a tricyclic antidepressant cannot be considered an inert substance. Second, individual sensitivity to drugs varies greatly, and even the first dose might have helped the patient paradoxically as a side effect if it had, for example, made her sleep better.
“SWOTTIES” RATHER THAN PLACEBOS?
In another recent paper, Evans and Hungin described a fictional general practice consultation, involving Mr Smith, a patient with irritable bowel syndrome, and his general practitioner, Dr Jones.7 Their discussion concentrates on a placebo, “SAS”, a specific drug that has been developed for the treatment of irritable bowel syndrome, and “Swotties”, well-known proprietary chocolate sweets and “a harmless and enjoyable substance”. An important detail is that, in clinical trials, the effectiveness of SAS and a placebo has been the same.
After a lengthy and sometimes complex chain of arguments, Evans and Hungin ended up with two main conclusions. First:
… when both (a) a drug fails to out-perform placebo and (b) the condition in question is a functional illness with no demonstrable underlying pathology, then the action of the drug is not only no better than placebo, and it is also no different from it either.
… in the circumstances of the consultation described, it is striking that current governance deems it ethical for a practitioner to prescribe either a drug or a placebo, both of which appear to rely for their effectiveness on a measure of concealment on the part of the doctor, yet deems it unethical for a practitioner openly to prescribe a harmless and enjoyable substance which (in equivalent conditions of transparency and information) is likely to be no less effective than either drug or placebo and is also likely to be better-tolerated and cheaper than the drug.
Dr Jones also concluded at the end of the paper that “the three preparations on her desk—the pharmaceutical product SAS, the “official” placebo and the green-candied Swotties, are functionally equivalent”.
The fictional consultation is described clearly; there are, however, several problems in the argumentation. First, it cannot be argued that the action of the drug does not differ from that of the placebo if it has failed to outperform a placebo in clinical trials. Even so, the drug as such has some unwanted side effects, while placebos as such are, by definition, inert.
Second, the authors do not pay attention to the fundamental difference between the two contexts in which placebos are used or can be used—the clinical context and the clinical trial context. Their difference is obvious if we consider the main mechanism of the placebo effect, namely, expectations.11 In the clinical encounter, the expectations of both the doctor and the patient are usually high. In a clinical trial, the patients are told that after randomisation they may end up in a placebo group; they understand this alternative and give their informed consent. The aim of the clinical encounter is to promote the health of the patient, while the aim of clinical research is to promote medical science and the health of future patients. Sometimes the trial participants are helped and sometimes they are harmed. The results of a clinical trial cannot simply be transferred into clinical practice.
Third, it is not true that “current governance deems it ethical for a practitioner to prescribe either a drug or a placebo, both of which appear to rely for their effectiveness on a measure of concealment on the part of the doctor”. While the open prescription of placebos in some conditions is accepted, concealed prescription is not. For example, the American Medical Association is explicit in its policy: “Physicians may use placebos for diagnosis or treatment only if the patient is informed of and agrees to its use.”3
Fourth, it cannot be said that Swotties are functionally equivalent to a placebo and SAS. While it is true that the latter two have been similarly successful in trials, no trials have used Swotties.
Fifth, Evans and Hungin argue that the lack of transparency is “generally supposed to underlie successful use of placebo—namely, to enjoy the benefits of the power of suggestion, it seems that the doctor must conceal from the patient at least some of the pertinent facts about the substance prescribed”.7 This general supposition is, however, not true. Placebos have been successfully used openly in clinical trials.12 13 A placebo effect can be produced by suggestions, but other mechanisms may be far more important. Past effects of active treatments and cues that signal that an active medication or treatment has been given are such mechanisms.14
Deliberate use of placebos is a common and widely accepted practice. However, the concept of the placebo is understood in several different ways, many of which do not refer to inert substances or treatments. In addition, the essential difference between placebo use and placebo effect is often not understood.15 Open administration of placebos does not include deceiving. It is, however, hardly ever justified. Referring to the favourable effects of placebos in some clinical trials is a half-truth, since the main factor behind the good results has been the context, not the placebo treatment as such. So called n-of-one trials with placebos are a special case of clinical trials.5
Funding: The Finnish Cultural Foundation has supported my work.
Competing interests: None declared.
Provenance and Peer review: Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.